Fault diagnosis-based SDG transfer for zero-sample fault symptom
The traditional fault diagnosis models cannot achieve good fault diagnosis accuracy when a new unseen fault class appears in the test set, but there is no training sample of this fault in the training set. Therefore, studying the unseen cause-effect problem of fault symptoms is extremely challenging...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Universitas Ahmad Dahlan
2023-11-01
|
| Series: | IJAIN (International Journal of Advances in Intelligent Informatics) |
| Subjects: | |
| Online Access: | http://ijain.org/index.php/IJAIN/article/view/1434 |
| _version_ | 1797403425618526208 |
|---|---|
| author | Mengqin Yu Yi Shan Lee Junghui Chen |
| author_facet | Mengqin Yu Yi Shan Lee Junghui Chen |
| author_sort | Mengqin Yu |
| collection | DOAJ |
| description | The traditional fault diagnosis models cannot achieve good fault diagnosis accuracy when a new unseen fault class appears in the test set, but there is no training sample of this fault in the training set. Therefore, studying the unseen cause-effect problem of fault symptoms is extremely challenging. As various faults often occur in a chemical plant, it is necessary to perform fault causal-effect diagnosis to find the root cause of the fault. However, only some fault causal-effect data are always available to construct a reliable causal-effect diagnosis model. Another worst thing is that measurement noise often contaminates the collected data. The above problems are very common in industrial operations. However, past-developed data-driven approaches rarely include causal-effect relationships between variables, particularly in the zero-shot of causal-effect relationships. This would cause incorrect inference of seen faults and make it impossible to predict unseen faults. This study effectively combines zero-shot learning, conditional variational autoencoders (CVAE), and the signed directed graph (SDG) to solve the above problems. Specifically, the learning approach that determines the cause-effect of all the faults using SDG with physics knowledge to obtain the fault description. SDG is used to determine the attributes of the seen and unseen faults. Instead of the seen fault label space, attributes can easily create an unseen fault space from a seen fault space. After having the corresponding attribute spaces of the failure cause, some failure causes are learned in advance by a CVAE model from the available fault data. The advantage of the CVAE is that process variables are mapped into the latent space for dimension reduction and measurement noise deduction; the latent data can more accurately represent the actual behavior of the process. Then, with the extended space spanned by unseen attributes, the migration capabilities can predict the unseen causes of failure and infer the causes of the unseen failures. Finally, the feasibility of the proposed method is verified by the data collected from chemical reaction processes. |
| first_indexed | 2024-03-09T02:38:22Z |
| format | Article |
| id | doaj.art-67c30420464b4362918e4f9f5d59653e |
| institution | Directory Open Access Journal |
| issn | 2442-6571 2548-3161 |
| language | English |
| last_indexed | 2024-03-09T02:38:22Z |
| publishDate | 2023-11-01 |
| publisher | Universitas Ahmad Dahlan |
| record_format | Article |
| series | IJAIN (International Journal of Advances in Intelligent Informatics) |
| spelling | doaj.art-67c30420464b4362918e4f9f5d59653e2023-12-06T06:59:49ZengUniversitas Ahmad DahlanIJAIN (International Journal of Advances in Intelligent Informatics)2442-65712548-31612023-11-019355156410.26555/ijain.v9i3.1434275Fault diagnosis-based SDG transfer for zero-sample fault symptomMengqin Yu0Yi Shan Lee1Junghui Chen2Chung Yuan Christian UniversityChung Yuan Christian UniversityChung Yuan Christian UniversityThe traditional fault diagnosis models cannot achieve good fault diagnosis accuracy when a new unseen fault class appears in the test set, but there is no training sample of this fault in the training set. Therefore, studying the unseen cause-effect problem of fault symptoms is extremely challenging. As various faults often occur in a chemical plant, it is necessary to perform fault causal-effect diagnosis to find the root cause of the fault. However, only some fault causal-effect data are always available to construct a reliable causal-effect diagnosis model. Another worst thing is that measurement noise often contaminates the collected data. The above problems are very common in industrial operations. However, past-developed data-driven approaches rarely include causal-effect relationships between variables, particularly in the zero-shot of causal-effect relationships. This would cause incorrect inference of seen faults and make it impossible to predict unseen faults. This study effectively combines zero-shot learning, conditional variational autoencoders (CVAE), and the signed directed graph (SDG) to solve the above problems. Specifically, the learning approach that determines the cause-effect of all the faults using SDG with physics knowledge to obtain the fault description. SDG is used to determine the attributes of the seen and unseen faults. Instead of the seen fault label space, attributes can easily create an unseen fault space from a seen fault space. After having the corresponding attribute spaces of the failure cause, some failure causes are learned in advance by a CVAE model from the available fault data. The advantage of the CVAE is that process variables are mapped into the latent space for dimension reduction and measurement noise deduction; the latent data can more accurately represent the actual behavior of the process. Then, with the extended space spanned by unseen attributes, the migration capabilities can predict the unseen causes of failure and infer the causes of the unseen failures. Finally, the feasibility of the proposed method is verified by the data collected from chemical reaction processes.http://ijain.org/index.php/IJAIN/article/view/1434conditional variational autoencoderfault diagnosissigned directed graphzero-shot learning |
| spellingShingle | Mengqin Yu Yi Shan Lee Junghui Chen Fault diagnosis-based SDG transfer for zero-sample fault symptom IJAIN (International Journal of Advances in Intelligent Informatics) conditional variational autoencoder fault diagnosis signed directed graph zero-shot learning |
| title | Fault diagnosis-based SDG transfer for zero-sample fault symptom |
| title_full | Fault diagnosis-based SDG transfer for zero-sample fault symptom |
| title_fullStr | Fault diagnosis-based SDG transfer for zero-sample fault symptom |
| title_full_unstemmed | Fault diagnosis-based SDG transfer for zero-sample fault symptom |
| title_short | Fault diagnosis-based SDG transfer for zero-sample fault symptom |
| title_sort | fault diagnosis based sdg transfer for zero sample fault symptom |
| topic | conditional variational autoencoder fault diagnosis signed directed graph zero-shot learning |
| url | http://ijain.org/index.php/IJAIN/article/view/1434 |
| work_keys_str_mv | AT mengqinyu faultdiagnosisbasedsdgtransferforzerosamplefaultsymptom AT yishanlee faultdiagnosisbasedsdgtransferforzerosamplefaultsymptom AT junghuichen faultdiagnosisbasedsdgtransferforzerosamplefaultsymptom |